I make AI agents safe to ship.

Bring the architecture you're about to commit to: I'll stress-test it against the failure modes that bite in production: credentials, state, audit, scale. You leave with the top risks, and a written memo within 48 hours.

Why listen to me

• I build this infrastructure in the open. mintkey, a credential broker so your agents never hold a real secret, and mint-mory: typed, self-consolidating agent memory. Reference implementations, not slideware.
• Day job: solutions architect. I review production architectures for a living. The weekend sessions apply the same method to your agent stack, before an incident or an auditor does.
• The method is public. The Agent Production Checklist: 40 questions your architecture must answer before agents touch prod. Free, no email gate. The paid session applies the same method to your system.
• The memo is the deliverable. Every session ends in a written one-pager of risks and recommendations you can circulate to your team, not a recording you'll never rewatch.